Network Congestion Control Method Based On Qos Routing Model

In recent years, the Internet becomes day by day busy with the rapid growth of the network user. The communication service swift and violent growth make the backbone network more and more jams, meanwhile, various new services set a higher request to the network QoS. The congestion control in the guarantee network stability and enhances the network QoS aspect to play the very important role. This thesis mainly focuses on the research of network congestion method based on QoS routing model, the main work and content are as follows:(1) Based on the system analysis the reason of network congestion and network QoS, an idea is proposed to solve the network congestion control by using the QoS routing model. The network simulation software NS2 is analyzed and extended function, which has laid the foundation for the network modeling and the network congestion control simulation research.(2) QoS routing model is studied analytically, the network is decomposed into various router units and the mathematical model of various router units is established based on the computer network abstract chart G(V,E), then connects them serially, constitutes the IP network system mathematical model to establish the QoS routing model in this model foundation.(3) Proposed a method of FGAPID network congestion control based on route model, and the method use the gray-prediction carries on the router queue length forecast in advance to detect the network congestion. To accomplishing network congestion control effectly, the parameters online self-adapting was implemented by FGAPID algorithm. The simulation results show that the method is effective and reliable.(4) Proposed a method of IAPSO path planning based on QoS routing model by combining Immune Algorithm and Particle Swarm Optimization Algorithm. The method plans the path of network load, which makes network load balancing and resource consumption as aim function to avoid network congestion by balancing the load and minimizing network resource consumption to meeting multifold indexes of bandwidth, delay and cost. The simulation results show that the method is effective and reliable.